Environ. Sci. Technol. 2006, 40, 7175-7185

Occurrence of a New Generation of the formation of trihalonitromethanes. In addition to the † chlorinated furanones that have been measured previously, Disinfection Byproducts brominated furanonesswhich have seldom been analyzeds were detected, especially in high-bromide waters. The STUART W. KRASNER,*,‡ presence of bromide resulted in a shift to the formation HOWARD S. WEINBERG,§ of other bromine-containing DBPs not normally measured SUSAN D. RICHARDSON,| SALVADOR J. (e.g., brominated ketones, acetaldehydes, nitromethanes, PASTOR,‡ RUSSELL CHINN,‡ acetamides). Collectively, ∼30 and 39% of the TOX and total MICHAEL J. SCLIMENTI,‡ organic bromine, respectively, were accounted for (on a § GRETCHEN D. ONSTAD, AND median basis) by the sum of the measured halogenated DBPs. ALFRED D. THRUSTON, JR.| In addition, 28 new, previously unidentified DBPs were Metropolitan Water District of Southern California, detected. These included brominated and iodinated haloacids, 700 Moreno Avenue, La Verne, California 91750-3399, Department of Environmental Sciences and a brominated ketone, and chlorinated and iodinated Engineering, University of North Carolina, aldehydes. Chapel Hill, North Carolina 27599-7431, and National Exposure Research Laboratory, U.S. Environmental Protection Agency, 960 College Station Road, Athens, Georgia 30605 Introduction Approximately 600-700 disinfection byproducts (DBPs) have been reported in the literature for the major disinfectants used (chlorine, ozone, chlorine dioxide, chloramines) as well A survey of disinfection byproduct (DBP) occurrence in as their combinations (1-3). Of these DBPs, only a small the United States was conducted at 12 drinking water percentage has been quantified in drinking waters. DBP treatment plants. In addition to currently regulated DBPs, surveys in the United States in the 1980s and 1990s provided more than 50 DBPs that rated a high priority for potential data for assessing a new maximum contaminant level (MCL) toxicity were studied. These priority DBPs included iodinated for trihalomethanes (THMs) as well as to develop regulations for other DBPs. In 1985, the U.S. Environmental Protection trihalomethanes (THMs), other halomethanes, a nonregulated Agency (EPA) measured chlorination DBPs at 10 operating haloacid, haloacetonitriles, haloketones, halonitromethanes, utilities, utilizing both target compound and broad-screen haloaldehydes, halogenated furanones, haloamides, analyses (2). The halogenated compounds, cumulatively, and nonhalogenated carbonyls. The purpose of this study accounted for between 30 and 60% of the total organic was to obtain quantitative occurrence information for halogen (TOX) found in these samples. In 1988-1989, a study new DBPs (beyond those currently regulated and/or studied) of 35 U.S. utilities was conducted, which analyzed for 19 for prioritizing future health effects studies. An effort halogenated DBPs and two aldehydes (3). On a weight basis, was made to select plants treating water that was high in THMs were the largest class of DBPs detected; the second total organic carbon and/or bromide to enable the largest fraction was haloacetic acids (HAAs). In addition, Glaze detection of priority DBPs that contained bromine and/or and Weinberg studied the formation of ozonation DBPs at 10 North American utilities in 1990-1991 (4). This study iodine. THMs and haloacetic acids (HAAs) represented the demonstrated that aldehydes could be removed with bio- two major classes of halogenated DBPs formed on a filtration (5), and bromate formation could be minimized at weight basis. Haloacetaldehydes represented the third a lower ozonation pH (6). major class formed in many of the waters. In addition to In 1997-1998, 296 U.S. utilities operating a total of 500 obtaining quantitative occurrence data, important new plants conducted a DBP survey under the Information information was discovered or confirmed at full-scale plants Collection Rule (ICR) (7). This survey included measurements on the formation and control of DBPs with alternative for the 4 regulated THMs, 6-9 HAAs (5 are regulated), 4 disinfectants to chlorine. Although the use of alternative haloacetonitriles, 2 haloketones, trichloronitromethane (chlo-

Downloaded via UNIV OF SOUTHERN CALIFORNIA on October 15, 2018 at 17:55:34 (UTC). disinfectants (ozone, chlorine dioxide, and chloramines) ropicrin), trichloroacetaldehyde (chloral hydrate), cyanogen See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. minimized the formation of the four regulated THMs, chloride, chlorite, chlorate, bromate, glyoxal, methyl glyoxal, trihalogenated HAAs, and total organic halogen (TOX), and 11 other aldehydes. The ICR, which included the same DBPs from the earlier studies (3, 4), greatly expanded our several priority DBPs were formed at higher levels with knowledge on the occurrence of these DBPs. the alternative disinfectants as compared with chlorine. For Other DBPs of health concern have had less extensive s example, the highest levels of iodinated THMs which monitoring. The chlorinated furanone 3-chloro-4-(dichlo- are not part of the four regulated THMsswere found at a romethyl)-5-hydroxy-2-(5H)-furanone (MX) has been mea- plant that used chloramination with no prechlorination. sured in a limited number of studies in the United States (8, The highest concentration of dichloroacetaldehyde was at 9) and elsewhere (10). For example, Kronberg and colleagues a plant that used chloramines and ozone; however, this found from 15 to 67 ng/L of MX in chlorinated drinking water disinfection scheme reduced the formation of trichloro- from three towns in Finland (10). MX, its geometric isomer acetaldehyde. Preozonation was found to increase (E)-2-chloro-3-(dichloromethyl)-4-oxobutenoic acid (EMX), and their oxidized and reduced forms were found in U.S. † This article is part of the Emerging Contaminants Special Issue. waters (11), while MX and brominated analogues of MX * Corresponding author phone: (909)392-5083; fax: (909)392-5246; (BMXs) have been identified in Japanese drinking waters e-mail: [email protected]. ‡ Metropolitan Water District of Southern California. (12). § University of North Carolina. Trihalonitromethanes with bromine have been identified | U.S. Environmental Protection Agency. in bench-scale chlorination studies (13), and bromopicrin

10.1021/es060353j CCC: $33.50  2006 American Chemical Society VOL. 40, NO. 23, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7175 Published on Web 07/26/2006 was detected in pilot-plant studies after ozone treatment Experimental Methods (14). In addition to trihalonitromethanes, mono- and diha- Chemical Standards. A number of the priority DBPs were logenated nitromethanes were identified in pilot-scale studies synthesized for this study (Supporting Information) (22, 26). of chlorine and chloramines, with and without preozonation Since the project began, some of the standards have become (15, 16). Brominated trihaloacetaldehydes have been found commercially available (Orchid Helix, New Westminster, BC, in chlorinated fulvic acid solutions containing bromide (17), Canada). Other synthesized standards used to confirm and iodinated THMs have been reported in chlorinated and broadscreen identifications were provided by CanSyn Chem chloraminated drinking water (18-21). Iodinated THM Corp. (Toronto, ON, Canada). All other chemicals and formation in bench- and plant-scale studies was favored by reagents were purchased at the highest level of purity from chloramination, especially if the ammonia was added first. Acros Organics (Pittsburgh, PA), Aldrich Chemical Co. Because many of the non-ICR DBPs have only been (Milwaukee, WI), ChemService (West Chester, PA), Fluka studied at bench scale or in limited full-scale surveys, there Chemical Co. (Ronkonkoma, NY), Mallinckrodt (Phillipsburg, is significant uncertainty over the identity and levels of NJ), Sigma Chemical Company (St. Louis, MO), Supelco DBPs that people are exposed to in their drinking water. (Bellefonte, PA), TCI America (Portland, OR), or Ultra Moreover, only a limited number of DBPs have been studied Scientific (North Kingston, RI) (23, 26). for adverse health effects because such studies are extremely expensive. Analytical Methods. A brief description of the analytical methods follows; more extensive details can be found in the To focus this research, an expert toxicology review of the Supporting Information. Most of the halogenated DBPss ∼500 DBPs reported in the literature (as of 1998) was THMs, tribromochloromethane, haloacetonitriles, haloke- conducted (22) with an in-depth mechanism-based structural tones, di- and trihalogenated acetaldehydes, halonitrometh- activity relationship analysisssupplemented by an extensive anes, tetrabromochloroethane, and benzyl chlorideswere literature search for genotoxicity and other datasused to analyzed for and quantified using a liquid/liquid extraction rank the carcinogenic potential of these DBPs. Approximately (LLE)-gas chromatography (GC)/electron capture detection 50 DBPs that received the highest ranking for potential (ECD) method (23, 26). A purge-and-trap-GC/mass spec- toxicity, and that were not already included in the ICR, were trometry (MS) method was used to analyze for VOCs and selected for a new U.S. occurrence study (23). Highest in this certain volatile chemicals that have been reported as possible ranking were the bromonitromethanes, which recently have DBPs (mono- and dihalogenated methanes, carbon tetra- been shown to be more cytotoxic and genotoxic than their chloride, methylethyl ketone) and to provide MS confirmation chlorinated analogues or their halomethane counterparts of other volatile DBPs (23, 26). A solid-phase extraction (SPE) (16, 24). Using in vitro mammalian cell assays, the bro- method was used to provide MS confirmation of semivolatile monitromethanes were more cytotoxic and genotoxic than DBPs (23, 26). A solid-phase microextraction (SPME)-GC/ most of the regulated HAAs (16, 25). ECD method was used in lieu of the LLE method in the last The priority DBPs in this study included mono-, di-, tri-, sampling quarter of the study (27). and/or tetra-substituted species of halomethanes (including iodinated species), haloacetonitriles, haloketones, halo- Nonhalogenated carbonyl compounds and chloroacetal- acetaldehydes, halonitromethanes, haloamides, and halo- dehyde were derivatized with pentafluoro-benzylhydroxyl- genated furanones (Table 1). Because most of the priority amine (PFBHA), and the oxime products were extracted and DBPs were from chlorine or chloramine disinfection, a few analyzed by GC/ECD (23). A LLE-GC/ECD method was used additional ozone and chlorine dioxide DBPs (e.g., carbonyls) for quantifying bromochloromethyl acetate (23). Another that were not ranked as a high priority were also included LLE-GC/ECD method was used for quantifying the halo- for completeness (i.e., to provide more information on those acetamides (23). The halogenated furanones were extracted alternative disinfectants). In addition, methyl tert-butyl ether from water, derivatized with boron trifluoride in methanol, (MTBE) and methyl bromide (bromomethane), which are back extracted, and analyzed by GC/ECD (23, 28). volatile organic compounds (VOCs) but not DBPs, were The ICR HAAs were measured using acidic and salted included in the list of target analytes because they are LLE, derivatization with acidic methanol, and GC/ECD important source water pollutants, and their measurement analysis (29). Another haloacid, 3,3-dichloropropenoic acid, would provide valuable occurrence information. Carbon was analyzed by a similar method, substituting diazomethane tetrachloride and methylethyl ketone, which are also VOCs, for acidic methanol in the derivatization step (30). TOC are possible DBPs. Regulated and some ICR DBPs were also measurements were made with the UV-persulfate oxidation included in this study for comparison purposes (Table 2). In method (31). The UVA measurements were made at 254 nm - addition, routine water quality measurements, such as total with filtered samples at ambient pH, using a UV visible organic carbon (TOC), ultraviolet absorbance (UVA) at 254 spectrophotometer. Bromide was measured by ion chro- nm, bromide, and TOX, were determined. matography (IC) (31). TOX was analyzed using the adsorption- The objectives of this new nationwide occurrence study pyrolysis titrimetric method (31). During the last sampling included the development of quantitative analytical methods quarter of the study, selected samples were also analyzed for for measuring the priority DBPs; the occurrence levels of total organic bromine (TOBr) and total organic chlorine these DBPsstogether with those in the ICRsin drinking (TOCl). The off-gas from the TOX combustion furnace was waters across the United States (including waters treated collected in water, and the concentrations of the bromide with chlorine, chloramines, ozone, and/or chlorine dioxide); and chloride ions were determined by IC (32). the effect of source water and treatment conditions on their Finally, selected samples were comprehensively (broad- formation; the fate and transport of these DBPs in the screen) analyzed for DBPs (to enable the detection of DBPs distribution system; and the identification of new, previously that were not among the ICR compounds and the group of unreported DBPs. Because chemical standards were often >50 priority DBPs quantified). GC with high- and low- not commercially available, several of the DBP standards resolution electron ionization (EI) and chemical ionization had to be synthesized for the methods development and (CI) MS was used primarily to identify the DBPs (23, 33). quantitation effort. This paper presents an overview of the When possible, tentative identifications were confirmed U.S. occurrence of a new generation of DBPs of health and through the analysis of authentic chemical standards, regulatory concern as well as water quality and treatment/ purchased commercially or synthesized. disinfection parameters that affect the formation and control Sampling Survey. A survey of 12 U.S. full-scale treatment of these DBPs. plants was conducted in the years 2000-2002. Each plant

7176 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 23, 2006 TABLE 1. Priority DBPs Selected for Nationwide Occurrence Study halomethanes chloromethane dichloroiodomethane bromodiiodomethaneb bromomethane (methyl bromide)a bromochloroiodomethane triiodomethane (iodoform)b bromochloromethane dibromoiodomethaneb carbon tetrachloride dibromomethane chlorodiiodomethaneb tribromochloromethane haloacids 3,3-dichloropropenoic acid haloacetonitriles chloroacetonitrile bromodichloroacetonitrile tribromoacetonitrile bromoacetonitrile dibromochloroacetonitrile haloacetates bromochloromethyl acetate haloketones chloropropanone 1-bromo-1,1-dichloropropanone 1,1,3,3-tetrabromopropanoneb 1,3-dichloropropanone 1,1,3,3-tetrachloropropanone 1,1,1,3,3-pentachloropropanonec 1,1-dibromopropanone 1,1,1,3-tetrachloropropanone hexachloropropanonec 1,1,3-trichloropropanone haloaldehydes chloroacetaldehyde bromochloroacetaldehydeb tribromoacetaldehydeb dichloroacetaldehyde halonitromethanes chloronitromethaneb bromochloronitromethaneb dibromochloronitromethaneb bromonitromethane dibromonitromethane tribromonitromethane (bromopicrin)b dichloronitromethaneb bromodichloronitromethaneb haloamides monochloroacetamideb dichloroacetamide trichloroacetamideb monobromoacetamideb dibromoacetamideb halogenated furanones 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone (MX) 3-chloro-4-(bromochloromethyl)-5-hydroxy-2(5H)-furanone (BMX-1) 3-chloro-4-(dichloromethyl)-2-(5H)-furanone (red-MX) 3-chloro-4-(dibromomethyl)-5-hydroxy-2(5H)-furanone (BMX-2) (E)-2-chloro-3-(dichloromethyl)butenedioic acid (ox-MX) 3-bromo-4-(dibromomethyl)-5-hydroxy-2(5H)-furanone (BMX-3) (E)-2-chloro-3-(dichloromethyl)-4-oxobutenoic acid (EMX) (E)-2-chloro-3-(bromochloromethyl)-4-oxobutenoic acid (BEMX-1)b (Z)-2-chloro-3-(dichloromethyl)-4-oxobutenoic acid (ZMX)g (E)-2-chloro-3-(dibromomethyl)-4-oxobutenoic acid (BEMX-2)b 2,3-dichloro-4-oxobutenoic acid (mucochloric acid) (MCA) (E)-2-bromo-3-(dibromomethyl)-4-oxobutenoic acid (BEMX-3)b (ring and open forms) VOCsd and miscellaneous DBPs 1,1,1,2-tetrabromo-2-chloroethanee methyl-tert-butyl ethera benzyl chloride 1,1,2,2-tetrabromo-2-chloroethaneb carbonyls 2-hexenal cyanoformaldehyde 6-hydroxy-2-hexanonef 5-keto-1-hexanalf methylethyl ketonef dimethylglyoxal a Not a DBP but included because it is an important source water contaminant. b DBP not originally prioritized (identified in drinking water after initial prioritization) but included due to similarity to other priority compounds. c Not analyzed; not stable in water. d Carbon tetrachloride, which is also a VOC, listed under halomethanes as a possible DBP. e Not analyzed; standard not available. f DBP not given a high priority but included for completeness sake to provide more representation to ozone DBPs for occurrence. g Isomer of EMX. was sampled four to five times to obtain information on Two plants that treat water from the same watershed but seasonal effects. Samples were collected in the fall (October- use different treatment/disinfection scenarios were sampled December) of 2000; the winter (plus early spring) (January- together (Table 3). An effort was made to select plants treating April), summer (July-September), and fall of 2001; and the water that was high in TOC and/or bromide to enable the winter (plus early spring) of 2002. Most of the halogenated detection of priority DBPs that contained bromine and/or DBPs were measured each quarter, whereas some analytical iodine (iodide levels were not measured, but in one study of fractions (i.e., certain carbonyls, haloacetamides, halogenated waters impacted by saltwater intrusion or connate water they furanones, 3,3-dichloropropenoic acid, and TOX) were were 8-25% of the bromide levels on a weight basis (34)). monitored for at half of the utilities each quarter (i.e., the All of the plants treated surface-water supplies except for latter parameters were measured at each utility up to 2-3 two plants (7 and 8), which treated a colored groundwater. times over the course of the study). Qualitative, broadscreen This study was not an occurrence survey per se, rather it was analyses were also conducted at half of the utilities each a targeted survey of waters with “challenged” conditions. quarter, with each plant sampled at least once over the course The treatment technologies encompassed by the plants of the study. included coagulation, lime softening, membrane softening,

VOL. 40, NO. 23, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7177 percentile values of 1.7 and 4.0 µg/L, respectively, in the TABLE 2. ICR and Regulated DBPs Included for Comparison ICR). The occurrence of chloral hydrate in the surveyed plants halomethanes (Table 4) was similar to that found in the ICR. With the addition of mono- and dihalogenated as well as bromine- chloroform dibromochloromethane containing species the haloacetaldehydes represented the bromodichloromethane bromoform third largest class of halogenated DBPs formed (on a weight haloacetonitriles basissalbeit, at much lower concentrations than that of the THMs or HAAs) in many of the waters in this study (Figure dichloroacetonitrile dibromoacetonitrile 1, Supporting Information). bromochloroacetonitrile trichloroacetonitrile In the plant effluents, the TOX was (on a median basis) haloacetic acids 178 µg/L as Cl-. In the ICR, the median value for surface- water plants was 122 µg/L. On a median basis, THM4, HAAs, monochloroacetic acid trichloroacetic acid monobromoacetic acid bromodichloroacetic acid and haloacetaldehydes accounted for 14, 12, and 2% of the dichloroacetic acid dibromochloroacetic acid TOX, respectively, in the surveyed plants (Figure 2, Supporting bromochloroacetic acid tribromoacetic acid Information). Collectively, ∼30% of the TOX was accounted dibromoacetic acid for (on a median basis) by the sum of the halogenated DBPs (halo-DBPs) (i.e., ∼70% remains unknown). For individual halonitromethanes plants, the percentage of TOX accounted for by the halo- trichloronitromethane (chloropicrin) DBPs ranged from 10 to 66% (Figure 2), which is similar to that reported in other studies (2). haloketones For seven plants sampled in the last quarter of the study, 1,1-dichloropropanone 1,1,1-trichloropropanone the effluent TOBr was (on a median basis) 79 µg/L as Br-, which is equivalent to 35 µg/L as Cl-. The percentage of haloaldehydes TOBr accounted for by the halo-DBPs ranged from 6 to 58%, chloral hydrate with a median value of 39% (Figure 3). This compares oxyhalides favorably to the range of values observed in a bench-scale study of disinfection of Suwannee river fulvic acid (35). In bromate chlorate that study, bromine-containing DBPs represented >60, 26, chlorite 14, and 8.2% of the TOBr in the samples treated with chlorine, chlorine dioxide, chloramines, and ozone, respectively. biological filtration, chlorination, ozonation, chlorine dioxide In the plant effluents, the TOCl was (on a median basis) disinfection, and chloramination. Most of the plants that 161 µg/L as Cl-. On a median basis, more of the TOX (as Cl-) used biological filtration had granular activated carbon (GAC) was due to TOCl than as TOBr. The percentage of TOCl filters or contactors. In addition, some plants added powdered accounted for by the halo-DBPs ranged from 12 to 33%, with activated carbon (PAC). a median value of 24% (Figure 3, Supporting Information). Zhang and colleagues found that the chlorine-containing Results and Discussion ICR DBPs represented ∼50, 31, 17, and 0% of the TOCl in the Overview. The concentrations of selected DBPs and their Suwannee river fulvic acid samples treated with chlorine, precursors are presented in Table 4. On a median basis, the chlorine dioxide, chloramines, and ozone, respectively (35). raw-water levels of TOC (5.8 mg/L) and bromide (0.12 mg/L) The percentage of TOX, TOBr, or TOCl that was accounted for the surveyed plants were much higher than those of the for in chlorinated waters tended to be higher than in samples plants studied in the ICR (median TOC and bromide were treated with alternative disinfectants (35). 2.4 and 0.04 mg/L, respectively). The raw-water specific UVA In the drinking water plant effluents, bromide utilization (SUVA), which is an indicator of the humic content of the in the halo-DBPs ranged from 1 to 41% of the raw-water water and provides an indication of the reactivity of TOC to bromide, with a median value of 22%. Amy and colleagues form DBPs (e.g., THMs, TOX), was (on a median basis) 2.9 found that when bromide is oxidized during disinfection to L/mg‚m for the surveyed plants. The median value corre- hypobromous acid, it is an efficient halogen substitution sponds to a source water of intermediate humic content. agent, and as much as 50% or more of the bromide was Figure 1 shows the occurrence (using box-and-whisker incorporated into THMs in formation potential tests (36). In plots) of the target DBPs and TOX in the finished waters of our full-scale study, in which chlorination doses were not as the 12 plants over the five quarters, and the Supporting high and reaction times not as long as in formation potential Information includes additional data for the 12 plants. As tests, somewhat less bromide utilization was observed and has been observed before (3, 7), THMs and HAAs were the was found to be incorporated into a wide range of DBPs, not two major classes of halogenated DBPs. Although the just THMs. surveyed plants treated waters relatively high in DBP precursors, the use of alternative disinfectants at most of the Occurrence of Different Classes of DBPs plants allowed them to minimize the formation of the sum THMs. The concentration of the iodinated THMs was typically of the four regulated THMs (THM4) in the plant effluents low compared to THM4 (Figure 1; Table 4). The ratio of the (median value of 31 µg/L). The median concentration of iodo-THMs to THM4 was 2% on a median basis, with 75th THM4 in finished waters in the ICRswhich included many percentile and maximum values of 7 and 81%, respectively. low-TOC groundwater systemsswas 23 µg/L. In the ICR, The highest formation of iodo-THMs was at plant 12 in when segmented by source of supply, the median THM4 November 2001 (iodo-THMs at 81% of the THM4), which value for distribution systems was 39 µg/L for surface-water added chlorine and ammonia simultaneously to form systems and 7.8 µg/L for groundwater systems. The sum of chloramines in a water with a moderate amount of bromide the nine HAAs in the surveyed plants was (on a median basis) (0.15 µg/L in November 2001) (Figure 4). There was no 34 µg/L, whereas the median value in the ICR was 20 µg/L. measurable free chlorine in this sample. As has been observed (ICR HAA data segmented by source of supply were not in previous research (19, 21), the formation of iodo-THMs available for comparison.) was favored by chloramination, especially if the ammonia In the ICR (7) and other studies (3), chloral hydrate was was added first. Overall, dichloroiodomethane was the most the only haloacetaldehyde measured (median and 75th common iodo-THM observed and was even found in waters

7178 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 23, 2006 TABLE 3. Treatment/Disinfection Processes at Surveyed Plants plant EPA no.a regionb treatmentc disinfectants 1 9 flocculation, coagulation, sedimentation, biological filtration ozone, chlorine, chloramines 2 9 coagulation, flocculation, sedimentation, filtration chlorine, chloramines 3 3 flocculation, coagulation, sedimentation, filtration, chlorine, chloramines biological filtration 4 3 flocculation, coagulation, sedimentation, filtration chlorine 5 4 flocculation, coagulation, sedimentation or solids ozone, chlorine contact-upflow clarification, biological filtration 6 4 coagulation, clarification, filtration chlorine dioxide, chlorine, chloramines 7 4 lime softening, filtration chloramines, ozone 8 4 lime softening, filtration or membrane softening chlorine, chloramines 9 7 lime softening, coagulation, sedimentation, PAC, filtration chlorine, chloramines 10 5 PAC, flocculation, coagulation, sedimentation, filtration chlorine, chloramines, or chlorine only or biological filtration 11 6 upflow-solids contact, flocculation/clarification, filtration chlorine dioxide, chlorine, chloramines 12 6 coagulation, filtration (chlorine dioxide), chloramines a The following pairs of plants treated water from the same or similar watersheds: plants 1 and 2; 3 and 4; 5 and 6; 7 and 8; 9 and 10; and 11 and 12. b The 12 plants in this survey were located in six of the nine regions defined by the USEPA. The states included in each of these six regions are as follows: EPA Region 9sArizona, California, Hawaii, Nevada; EPA Region 3sDelaware, Maryland, Pennsylvania, Virginia, West Virginia, Washington D.C.; EPA Region 4sAlabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, Tennessee; EPA Region 7sIowa, Kansas, Missouri, Nebraska; EPA Region 5sIllinois, Indiana, Michigan, Minnesota, Ohio, Wisconsin; EPA Region 6sArkansas, Louisiana, New Mexico, Oklahoma, Texas. c Some plants had parallel trains with different treatment processes.

TABLE 4. Concentration of Selected DBPs and Their Precursors in U.S. Occurrence Study in Waters Exhibiting High-Precursors Loadingsa parameter unit location minimum median 75th percentile maximum TOC mg/L raw water 3.0 5.8 13 SUVA L/mg-m raw water 1.9 2.9 3.9 bromide mg/L raw water 0.02 0.12 0.40 THM4 µg/L plant effluent 4 31 45 164 sum of 6 iodinated THMs µg/L plant effluent NDb 0.4 2 19 sum of 9 HAAs µg/L plant effluent 5 34 56 130 3,3-dichloropropenoic acid µg/L plant effluent NDb (<0.1) NDb 0.7 4.7 sum of haloacetaldehydes µg/L plant effluent 0.2 4 7 20 chloral hydrate µg/L plant effluent NDb 13 16 dichloroacetaldehyde µg/L plant effluent NDb 12 14 sum of haloacetonitriles µg/L plant effluent NDb 34 14 dichloroacetonitrile µg/L plant effluent NDb 12 12 sum of haloacetamides µg/L plant effluent NDb 1.4 2.5 7.4 2,2-dichloroacetamide µg/L plant effluent NDb 1.3 2.0 5.6 sum of haloketones µg/L plant effluent NDb 24 9 1,1,1-trichloropropanone µg/L plant effluent NDb 0.8 3 7 1-bromo-1,1-dichloropropanone µg/L plant effluent NDb 0.2 <1 <3 sum of halonitromethanes µg/L plant effluent NDb 13 10 chloropicrin µg/L plant effluent NDb 0.2 0.4 2.0 bromopicrin µg/L plant effluent NDb NDb NDb 5 sum of halogenated furanones ng/L plant effluent NDb 310 610 2,380 MX ng/L plant effluent NDb (<20) 20 60 310 BMX-1 ng/L plant effluent NDb (<20) NDb 80 170 BEMX-1 ng/L plant effluent NDb (<20) NDb NDb 720 BMX-2 ng/L plant effluent NDb (<20) NDb NDb 30 BEMX-2 ng/L plant effluent NDb (<20) NDb 30 810 BMX-3 ng/L plant effluent NDb (<20) NDb NDb 40 BEMX-3 ng/L plant effluent NDb (<20) NDb 180 410 TOX µg/L as Cl- plant effluent 21 178 206 284 TOCl µg/L as Cl- plant effluent 87 161 194 206 TOBr µg/L as Br- plant effluent 36 79 80 229 TOBr µg/L as Cl- plant effluent 16 35 36 102 cyanoformaldehyde µg/L plant effluent NDb NDb NDb 0.3 5-keto-1-hexanal µg/L plant effluent NDb NDb NDb NDb 6-hydroxy-2-hexanone µg/L plant effluent NDb NDb NDb NDb dimethylglyoxal µg/L plant effluent NDb 0.1 1.4 3.5 trans-2-hexenal µg/L plant effluent NDb NDb NDb 0.7 methyl ethyl ketone µg/L plant effluent NDb NDb NDb 5 a Twelve plants, up to 5 seasons each. b ND ) not detected. that had average levels of bromide (which should also have Haloacids. The one priority haloacid that was quantified had some iodide) (e.g., note plants 5 and 6 in the Supporting in this study was 3,3-dichloropropenoic acid (Table 4). Its Information). determination provides quantitative evidence that haloge-

VOL. 40, NO. 23, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7179 FIGURE 1. Concentration of DBPs in plant effluents in U.S. occurrence study in waters exhibiting high-precursors loadings (12 plants, up to 5 seasons each). I-THMs ) iodinated THMs, HANs ) ) ) FIGURE 4. Effect of chloramination on iodo-THM formation at plant haloacetonitriles, HAs haloacetaldehydes, HKs haloketones, ) HNMs ) halonitromethanes, HAMs ) haloacetamides, HFs ) 12 effluent (11/15/01, bromide 0.15 mg/L). The x-axis shows halogenated furanones, and CdO ) carbonyls. decreasing chlorine (Cl) and increasing bromine/iodine substitution.

genated HAAs can be formed by chlorine dioxide (35, 37-39), chlorite and chlorate are generally the only chlorine dioxide DBPs considered in most chlorine dioxide treatment studies. In our study, DXAAs were detected after predisin- fection with chlorine dioxidese.g., 16 µg/L at plant 6 in November 2000sbut no trihalogenated HAAs were formed, and THM formation was quite low (2 µg/L) in this example (Figure 4, Supporting Information). After the addition of free chlorine, the levels of HAAs increased, including the forma- tion of trihalogenated species. At plant 12, a significant level of total HAAs (32 µg/L) was produced during pretreatment with chlorine dioxide disinfection (this predisinfectant was only used at this plant during the February 2002 sampling), whereas very little THMs (3 µg/L) were formed. The majority of the HAAs produced were DXAAs (21 µg/L). In this high- bromide water (0.33 mg/L), all three DXAAs were formed, FIGURE 2. Percentage of TOX accounted for in plant effluents in which included a significant amount of dibromoacetic acid. U.S. occurrence study (12 plants, up to 3 seasons each). I-THMs ) Many brominated acids were identified in several finished iodinated THMs, HANs ) haloacetonitriles, HAs ) haloacetalde- waters that contained elevated levels of bromide in their hydes, HKs ) haloketones, HNMs ) halonitromethanes, HAMs ) source waters. A number of them were identified for the first haloacetamides, and HFs ) halogenated furanones. time (Table 5), with carbon chain lengths of three and four being common as well as the presence of diacids and double bonds in their structures. Many of these products have also been recently identified in drinking waters from Israel containing elevated levels of bromide (∼2 mg/L) that have been treated with chlorine or chlorine dioxide-chloramines (39). The most unusual bromo acids identified in this study included the bromo-oxo-acids (3,3-dibromo-4-oxopentanoic acid and 3-bromo-3-chloro-4-oxopentanoic acid), which were identified in finished drinking waters from plant 1 and 11. 3,3-Dibromo-4-oxopentanoic acid and its chlorinated analogue, 3,3-dichloro-4-oxopentanoic acid, have recently been identified, respectively, in drinking water from Israel treated with chlorine or chlorine dioxide-chloramines (39) and in pilot-plant water that was either chlorinated or ozonated-chlorinated (40). The only other previous report of halo-oxo-acids was the tentative identification of 2,3- dichloro-4-oxopentanoic acid and trichloro-4-oxopentanoic FIGURE 3. Percentage of TOBr accounted for in plant effluents in acid in a previous bench-scale study involving the reaction U.S. occurrence study (7 plants, 1 season each). I-THMs ) iodinated ) ) of humic acids with chlorine (cited in ref 1). THMs, HANs haloacetonitriles, HAs haloacetaldehydes, HKs The EI mass spectrum for an unusual brominated oxo- ) haloketones, HNMs ) halonitromethanes, HAMs ) haloacet- acid identified is shown in Figure 5, Supporting Information. amides, and HFs ) halogenated furanones. A discussion of how this new DBP was identified is provided nated acids other than HAAs are DBPs. Also, it is significant in the Supporting Information. to note that certain haloacids (dihalogenated HAAs [DXAAs]) Another significant finding in this study was the discovery were formed upon treatment with chlorine dioxide. Although of iodoacids for the first time: iodoacetic acid, bromoiodoace- some previously published studies (bench-, pilot-, and full- tic acid, (E)-3-bromo-3-iodopropenoic acid, (Z)-3-bromo- scale) have indicated that DXAAs and sometimes trihalo- 3-iodopropenoic acid, and (E)-2-iodo-3-methylbutenedioic

7180 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 23, 2006 TABLE 5. New DBPs Identified at Surveyed Plants in Waters Exhibiting High-Precursors Loadingsa brominated haloacids 2,2-dibromopropanoic acid 3-bromo-3-chloro-4-oxopentanoic acid dibromochloropropanoic acid 3,3-dibromo-4-oxopentanoic acid 3,3-dibromopropenoic acid bromoheptanoic acid cis-2,3-dibromopropenoic acid bromochloroheptanoic acid tribromopropenoic acid dibromoheptanoic acid 2-bromobutanoic acid bromochlorononanoic acid trans-4-bromo-2-butenoic acid cis-2-bromobutenedioic acid cis-4-bromo-2-butenoic acid trans-2,3-dibromobutenedioic acid trans-2,3-dibromo-2-butenoic acid cis-2-bromo-3-methylbutenedioic acid bromodichlorobutenoic acid iodinated haloacids iodoacetic acid (Z)-3-bromo-3-iodopropenoic acid bromoiodoacetic acid (E)-2-iodo-3-methylbutenedioic acid (E)-3-bromo-3-iodopropenoic acid brominated haloketone 1-bromo-1,3,3-trichloropropanone halogenated aldehydes iodobutanal 4-chloro-2-butenal dichloropropenal a DBPs confirmed with authentic standards are underlined; all other DBP identifications should be considered tentative.

FIGURE 5. Halofuranones (MX and BMX analogues) in drinking water from EPA Region 6. FI ) filter influent; FE ) filter effluent; PE ) plant effluent; DS ) distribution system; SDS ) simulated distribution system. acid (Table 5). These iodoacids were found in finished bromo acetic acids) (41). Iodoacetic acid has also been shown drinking water from plant 12 in November 2001, which used to cause developmental effects in mouse embryos (neural chloramines only and also had relatively high levels of iodo- tube closures) at low µM levels (42). Recently, it was THMs. High-resolution MS confirmed the presence of iodine demonstrated that iodoacetic acid induced its genotoxic in the overall empirical formulas for these new DBPs. damage via an oxidative stress mechanism (43). The other Tentative identifications were confirmed through the analysis four iodoacids discovered are the subject of current toxicity of purchased (iodoacetic acid) and synthesized standards studies. (41). Although these iodoacids were not identified prior to This study demonstrates that in waters high in bromide the prioritization effort for the selection of high priority DBPs, (and iodide), more brominated and/or iodinated compounds they are likely to be toxicologically important. Recently (e.g., THMs, haloacids) can form. The ICR (and some previous conducted mammalian cell cytotoxicity and genotoxicity research) focused on one- and two-carbon structures. This studies have shown that iodoacetic acid is 2-fold more work demonstrates that analogues of previously identified genotoxic and 3-fold more cytotoxic than bromoacetic acid structures are found for longer carbon chain lengths (even (which is the most potent genotoxin/cytotoxin of the chloro/ up to nine carbons).

VOL. 40, NO. 23, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7181 FIGURE 6. Relative formation (and speciation) of haloacetamides and HAAs in plant 12 effluent (2/12/02, bromide ) 0.33 mg/L). Haloacetamides not measured in this study included bromochloro, bromodichloro, dibromochloro, and tribromo species.

Haloketones. The priority haloketone 1-bromo-1,1- previously identified, which were tentatively identified as dichloropropanone reached a maximum of 3 µg/L in a dichloropropenal and 4-chloro-2-butenal. distribution system sample for plant 5 in November 2000 Halonitromethanes. The priority halonitromethane bro- (which used ozone-chlorine disinfection). In addition to the mopicrin reached a maximum of 5 µg/L in the effluent of target haloketones, mixed bromochloro analogues were plant 12 in Feburary 2002 (which used chlorine dioxide- detected in selected samples by the broadscreen GC/MS chloramine disinfection). This sample also had the highest methods. For example, at plant 7 (where the raw-water halonitromethane sum (10 µg/L). In addition, preozonation bromide was 0.12 mg/L), six tetrahalogenated species were was found to increase the levels of selected haloni- detected after chloramination and ozonation (1,1,3,3-tetra- tromethanes at other plants. Figures 6 and 7, Supporting chloro-; 1-bromo-1,3,3-trichloro-; 1,1-dibromo-3,3-dichloro-; Information, illustrate this for a comparison of the same 1,3-dibromo-1,3-dichloro-; 1,1,3-tribromo-3-chloro-; and source water treated (8/13/01) with ozone-chlorine (plant 5) 1,1,3,3-tetrabromopropanone). One of these haloketones, or chlorine dioxide-chlorine-chloramines (plant 6) and 1-bromo-1,3,3-trichloropropanone, was identified in drinking another source water treated (July 2001) with ozone-chlorine- water for the first time and was found in many of the waters chloramines (plant 1) or chlorine-chloramines (plant 2). Other sampled, although the range of concentrations is unknown. research has shown that preozonation can increase the Haloaldehydes. The priority haloaldehyde dichloro- formation of chloropicrin (45) or other halonitromethanes acetaldehyde reached a maximum of 16 µg/L in a simulated (15, 16) upon postchlorination. distribution system sample for plant 7 in December 2000 Preozonation and moving the point of chlorination until (which used chloramine-ozone disinfection). Ozonation after coagulation at plant 1 (chlorine was added to the raw without biological filtration and chloramination was found water at plant 2) also caused a shift in speciation to the more to increase the formation of dihaloaldehydes. For example, brominated forms of the trihalonitromethanes; similar shifts chlorine-chloramines disinfection at plant 8 (12/11/00) were also observed for the THMs. At plants 1 and 2 (July produced 13 µg/L of chloral hydrate and 3 µg/L of dichlo- 2001), the THM bromine incorporation factors (molar sum roacetaldehyde, whereas ozone and chloramines at plant of bromine incorporated into THMs divided by molar sum 7swhich treated groundwater from the same aquifer as plant of THM4)swhere the bromine incorporation factor ranges 8sproduced 0.3 µg/L of chloral hydrate and 12 µg/L of from 0 (all chloroform) to 3 (all bromoform) (46)swere 1.8 dichloroacetaldehyde. McKnight and Reckhow found that and 1.0, respectively, whereas the bromine incorporation acetaldehyde (an ozone DBP) can react with chlorine to factors for trihalogenated nitromethanesswhere the bromine initially form chloroacetaldehyde, which in the presence of incorporation factor ranges from 0 (all chloropicrin) to 3 (all excess free chlorine, can rapidly react to form chloral hydrate bromopicrin)swere 2.3 and 1.3, respectively. The bromine (44). At plant 7, it is possible that with only a small free chlorine incorporation factor for each class of DBP was similar at residual (if any) that the conversion to the trichlorinated each plant; in other words, the shift in bromine speciation species was concentration-limited and the reaction termi- for each class of DBP was similar at each plant. The differ- nated with dichloroacetaldehyde. ence in speciation between the two plants (i.e., higher An iodinated DBPscharacterized by high-resolution MS bromine incorporation at plant 1) was explained (in part) as a product with the empirical formula C4H7OI and molecular by the higher bromide-to-TOC ratio at the point of chlorina- weight of 198stentatively identified as iodobutanal, was tion at plant 1 (36). In addition, the result was also related found in finished waters from two treatment plants (3 and to differences in chlorine doses. Plant 1 applied a lower 6). Plants 3 and 6 used chlorine-chloramine and chlorine chlorine dose (2.2 mg/L) than plant 2 (total chlorine dose ) dioxide-chlorine-chloramine disinfection. This iodoaldehyde 4.25 mg/L). Therefore, the chlorine/bromide ratio was was not detected at plant 12, which had relatively high levels lower at plant 1, resulting in a higher bromine incorpora- of iodo-THMs. This is the first report of an iodoaldehyde as tion (46). a DBP in drinking water. It is highly likely that the molecule Halogenated Furanones (MX Analogues). The highest is an iodoaldehyde with four carbons, but its exact structure levels for halogenated furanone sums occurred at a plant is not currently known. Broadscreen analyses also revealed (plant 8) that disinfected a water with 11.3 mg/L of TOC and the presence of two other haloaldehydes that have not been 0.27 mg/L of bromide with chlorine-chloramines (2380 ng/L

7182 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 23, 2006 in plant effluent drinking water, Supporting Information) the nonhalogenated priority carbonyls, dimethylglyoxal was and at a plant (plant 11) that disinfected a water with 3.5 consistently the most prevalent. mg/L of TOC and 0.21 mg/L of bromide with chlorine dioxide- chlorine-chloramines (1000 ng/L in the distribution system, Impact on Drinking Water Treatment and Future Health Figure 5). Because many of the source waters in this study Effects Studies were relatively high in TOC and/or bromide, high occurrence of some of the halogenated furanones was observed at The quantitative information provided through this study selected plants (Figure 8, Supporting Information). MX levels was critical for prioritizing future health effects research. >100 ng/L were detected in 7 of 42 filter effluent, plant Toxicity studies continue for the newly identified iodoacids, effluent, and distribution system samples. MX levels reached and in vivo studies are currently being conducted for some a high of 850 ng/L in the average detention time distribution of the bromonitromethanes. New toxicity studies are also system sample from plant 11 on 9/10/01. The median and underway for haloamides and haloacetonitriles (51) as well 75th percentile plant effluent levels of MX were similar to the as the iodo-THMs. It is likely that other priority DBPs will be levels found in other studies (8-10). However, the maximum studied for potential adverse health effects. Because the occurrence was high compared to previous limited surveys magnitude of the risk of a particular chemical includes both (e.g., 80 ng/L in two U.S. studies (8, 9), 260 ng/L in a its concentration and toxic potency, assuming a health Californian study (47), 90 ng/L in an Australian study (48), priority for specific DBPs from this study without the health and 586 ng/L in a Russian study (49)). effects data would be premature. For this reason, concentra- The results of this survey indicate that BMXs and tions shown in this paper should be considered along with brominated EMXs (BEMXs) can occur at relatively high health effects data. For example, brominated and iodinated concentrations in some high-bromide waters (Table 4). DBPs are much more potent than their chlorine-containing However, the median plant effluent levels of the six BMX counterparts, which may be because bromine and iodine and BEMX species was each <20 ng/L. In a study in Japan are better leaving groups than chlorine (22). Thus, DBPs such (12), the maximum occurrence of BMX-1, BMX-2, and BMX-3 as the brominated MX analogues, bromonitromethanes or was 2, 6, and 11 ng/L, respectively, but the maximum MX iodo-THMs should not be dismissed because they are present formation was also relatively low (6 ng/L) compared to this at lower levels than some of the other DBPs, because their study or to other surveys (8-10, 47-49). potency could be much greater than other DBPs present at The halogenated furanones were often stable in the higher concentrations. Conversely, DBPs present at higher distribution systems and in simulated distribution system concentrations should not necessarily be assumed to be the tests, most of which were in chloraminated water. Previous most important until their toxic potencies are known. controlled laboratory studies had suggested that MX may This study also has important implications for drinking not be stable in chlorinated distribution systems (8). In two water treatment. Because the commonly used alternative instances, MX levels increased in concentration from the disinfectants (ozone, chloramines, and chlorine dioxide) plant effluent to the distribution system point sampled (for produce lower levels of the four regulated THMs and most plant 6 during one sampling event and for plant 11 during of the HAAs as well as TOX, many water utilities have switched one sampling event). In another instance, MX levels decreased (or are in the process of switching) from chlorine to these in the distribution system (for plant 3). alternative disinfectants to meet the Stage 1 and/or Stage 2 Haloamides. Several haloacetamides were quantified for DBP Rules. Although the four regulated THMs and most HAAs the first time in this study (Figure 1, Supporting Information) are minimized with the use of alternative disinfectants, some and found to be present at levels similar to other commonly of the priority DBPs were higher in concentration with the measured DBPs (Table 4). The highest level for the halo- use of alternative disinfectants. For example, iodo-THM levels acetamide sum (14 µg/L) occurred in a simulated distribution were highest in a water disinfected with chloramines only; system sample for plant 6 that disinfected a water with 9.5 dichloroacetaldehyde was highest in a water disinfected with mg/L of TOC and 0.06 mg/L of bromide with chlorine dioxide- chloramines and ozone (with no biological filtration); selected chlorine-chloramines. Alkaline hydrolysis of haloacetonitriles halonitromethanes were formed at higher levels during can form haloacetamides (and ultimately HAAs) (50). For postdisinfection when preozonation was used; and MX example, dichloroacetonitrile can hydrolyze to form dichlo- precursors were better controlled at plants using ozone than roacetamide, which can further hydrolyze to form dichlo- those using chlorine dioxide. Because disinfection with roacetic acid. Therefore, it is possible that some or all of the chloramines only may not always meet virus or Giardia haloacetamides were formed by the hydrolysis of halo- inactivation requirements, this disinfection scenario is acetonitriles. The occurrence of 2,2-dichloroacetamide was not used at many plants that treat surface water. When similar to that of dichloroacetonitrile (Table 4). Figure 6 shows chloramines are used as the secondary disinfectant at a the relative formation of haloacetamides and HAAs at plant drinking water treatment plant, it should be possible to limit 12 (2/12/02, bromide ) 0.33 mg/L). Although not all of the the formation of iodo-DBPs through a sufficient free-chlorine bromine- and chlorine-containing haloacetamide species contact time before the addition of ammonia to form were analyzed in this study, it does appear that the relative chloramines or through the use of preozonation (19, 21). For concentrations of the haloacetamides compared well to the example, in this study, iodinated THM formation was low at HAAs (i.e., similar degree of halogen incorporation), where plants using free chlorine or ozone for primary disinfection. the haloacetamides were in a concentration range of ∼10 In addition, it should be possible to minimize the formation times less than the HAAs. The dichloro species of each DBP of haloaldehydes at ozone plants through the use of biological class were formed at higher concentrations than the trichloro filtration. One of these haloaldehydes, dichloroacetaldehyde, species, where the presence of an elevated level of bromide is of moderate health concern, because it is a potential cross- resulted in a shift in speciation of the dihalogenated species linking agent (22). to the more bromine-substituted compounds. Thus, it is likely For some waters with high TOC and/or high bromide, that brominated analogues of trichloroacetamide were also relatively high MX and BMX levels were detected in this study. formed in this water. Haloamides are a new family of DBPs This new occurrence information should be considered with of health concern (51). existing health effects information to address the potential Nonhalogenated Carbonyls. Several nonhalogenated risk of these DBPs at those plants. Also, higher-chain haloacids carbonyls were quantified in drinking water for the first time (such as the 3,3-dichloropropenoic acid) should be consid- in this study (Figure 1, Supporting Information; Table 4). Of ered along with the commonly measured HAAs. It is

VOL. 40, NO. 23, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7183 important to acknowledge that three- and four-carbon pentanoic acid methyl ester; effect of preozonation on haloacids can be formed in drinking water treatmentsnot halonitromethane formation at two plants; concentration of only the two-carbon HAAs that are regulated. Finally, the halogenated furanones in plant effluents; DBPs measured new information on haloacetamide concentrations is im- and identified in finished water at drinking water treatment portant because this is the first quantitative data for this plants; list of synthesized standards; selected analytical DBP class, and the data demonstrate that their levels can be methods; and discussion of how a new DBP was identified. comparable to other commonly measured DBPs. Health This material is available free of charge via the Internet at effects information is needed for the haloacetamides as well http://pubs.acs.org. as other DBPs identified in this survey. Since this project was completed, many of the priority Literature Cited and newly identified DBPs have been incorporated into new (1) Richardson, S. D. Drinking water disinfection by-products. In studies. However, these compounds are not the end of the Encyclopedia of Environmental Analysis and Remediation; story. Emerging DBPs of health and regulatory concern (e.g., Meyers, R. A., Ed.; John Wiley & Sons: New York, 1998; Vol. 3, nitrosamines) are being identified in disinfected drinking pp 1398-1421. water. Because all chemical disinfection processes produce (2) Stevens, A. A.; Moore, L. A.; Slocum, C. J.; Smith, B. L.; Seeger, DBPs, it is important to include a wide range of potential D. R.; Ireland, J. C. By-products of chlorination at ten operating DBPs in treatment studies to determine how best to minimize utilities. In Water Chlorination: Chemistry, Environmental Impact and Health Effects; Jolley, R. L., Condie, L. W., Johnson, the formation of as many DBPs as possible, recognizing that J. D., Katz, S., Minear, R. A., Mattice, J. S., Jacobs, V. A., Eds.; not all will be minimized (and some may be maximized). Lewis Publishers: Chelsea, MI, 1990; Vol. 6, pp 579-604. (3) Krasner, S. W.; McGuire, M. J.; Jacangelo, J. G.; Patania, N. L.; Acknowledgments Reagan, K. M.; Aieta, E. M. The occurrence of disinfection by- products in U.S. drinking water. J. Am. Water Works Assoc. 1989, The U.S. EPA funded and collaborated in the research 81 (8), 41-53. described here. Acknowledgment is given to other chemists (4) Glaze, W. H.; Weinberg, H. S. Identification and Occurrence of and laboratory technologists who helped develop the ana- Ozonation By-Products in Drinking Water; American Water Works Association Research Foundation (AWWARF) and Ameri- lytical methods, synthesized standards, and conducted or can Water Works Association (AWWA): Denver, CO, 1993. supervised the analyses at Metropolitan Water District of (5) Weinberg, H. S.; Glaze, W. H.; Krasner, S. W.; Sclimenti, M. J. Southern California (Alicia Gonzalez, Lely Suhady, Jacob Formation and removal of aldehydes in plants that use Nikonchuk, Leslie Bender, Tim Albrecht, Vaheh Martyr, Sylvia ozonation. J. Am. Water Works Assoc. 1993, 85 (5), 72-85. Barrett, Hsiao-Chiu Wang, Suzanne Teague, Ching Kuo, (6) Krasner, S. W.; Glaze, W. H.; Weinberg, H. S.; Daniel, P. A.; Najm, Robert Alvarez, Jesus Vasquez, Jr., Bart Koch, Eric Crofts, I. N. Formation and control of bromate during ozonation of waters containing bromide. J. Am. Water Works Assoc. 1993, 85 Sikha Kundu, Tiffany Lee, Pat Hacker, and Himansu Mehta), (1), 73-81. University of North Carolina (UNC) (Ramiah Sangaiah, (7) McGuire, M. J.; McLain, J. L.; Obolensky, A. Information Karupiah Jayaraj, Christine N. Dalton, Lindsay Dubbs, Gary Collection Rule Data Analysis; AwwaRF and AWWA: Denver, L. Glish, Katrina Jamison, Vanessa Pereira, Petra Strunk, and CO, 2002. Zhengqi Ye) and the U.S. EPA (Terrance L. Floyd and F. Gene (8) Meier, J. R.; Knohl, R. B.; Coleman, W. E.; Ringhand, H. P.; Munch, Crumley). We express appreciation to Leif Kronberg (A° bo J. W.; Kaylor, W. H.; Streicher, R. P.; Kopfler, F. C. Studies on the potent bacterial mutagen, 3-chloro-4-(dichloromethyl)-5- Akademi, Finland) and Angel Messeguer (Institute of Chemi- hydroxy-2(5H)-furanone: aqueous stability, XAD recovery and cal and Environmental Research, Barcelona, Spain) for analytical determination in drinking water and chlorinated providing samples of furanone standards as well as Bruce humic acid solutions. Mutat. Res. 1987, 189, 363-373. McKague (CanSyn Chem Corp., Toronto, Canada), Francesc (9) Wright, J. M.; Schwartz, J.; Vartiainen, T.; Ma¨ki-Paakkanen, J.; Ventura (Aigues of Barcelona, Spain), and George Majetich Altshul, L.; Harrington, J. J.; Dockery, D. W. 3-Chloro-4- (dichloromethyl)-5-hydroxy-2(5H)-furanone (MX) and mu- (Majestic Research, Athens, GA) for providing standards used tagenic activity in Massachusetts drinking water. Environ. Health in the quantitative methods and new DBP identification work. Perspect. 2002, 110 (2), 157-164. The authors acknowledge Philip Singer of UNC for helping (10) Kronberg, L.; Holmbom, B.; Reunanen, M.; Tikkanen, L. to select and solicit the participating utilities for this study Identification and quantification of the Ames mutagenic as well as for his assistance in developing the sampling plans compound 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-fura- for many of the plants. We gratefully acknowledge the none and of its geometric isomer (E)-2-chloro-3-(dichloro- methyl)-4-oxobutenoic acid in chlorine-treated humic water assistance of the participating utilities that collected the and drinking water extracts. Environ. Sci. Technol. 1988, 22 (9), samples, provided operational and water quality data, and 1097-1103. ensured a successful survey. We are also grateful to the U.S. (11) Kronberg, L.; Singh, R.; Ball, L.; Johnson, J. D.; Christman, R. F. EPA Office of Water and Office of Prevention, Pesticides, and Identification of Mutagenic By-Products from Aquatic Humic Toxic Substances scientistssYin-tak Woo, Jennifer McLain, Chlorination; AWWARF: Denver, CO, 1990. Vicki Dellarco, David Lai, and Mary Ko Manibusanswho (12) Suzuki, N.; Nakanishi, J. Brominated analogues of MX (3-chloro- 4-(dichloromethyl)-5-hydroxy-2(5H)-furanone) in chlorinated undertook the initial DBP prioritization effort that provided drinking water. Chemosphere 1995, 30 (8), 1557-1564. the focus for this study. We are especially grateful to Vicki (13) Thibaud, T.; De Laat, J.; Dore, M. Effects of bromide concentra- Dellarco for her vision in recognizing the importance of such tion on the production of chloropicrin during chlorination of a prioritization effort. This paper has been reviewed in surface waters. Formation of brominated trihalonitromethanes. - accordance with the U.S. EPA’s peer and administrative Water Res. 1988, 22 (3), 381 390. (14) Krasner, S. W.; Chinn, R.; Hwang, C. J.; Barrett, S. E. Analytical review policies and approved for publication. Mention of methods for brominated organic disinfection by-products, Pro- trade names or commercial products does not constitute ceedings of the 1990 AWWA Water Quality Technology Confer- endorsement or recommendation for use by the U.S. EPA. ence (WQTC); AWWA: Denver, CO, 1991. (15) Richardson, S. D.; Thruston, A. D., Jr.; Caughran, T. V.; Chen, Supporting Information Available P. H.; Collette, T. W.; Floyd, T. L.; Schenck, K. M.; Lykins, B. W., Jr.; Sun, G.-R.; Majetich, G. Identification of new drinking water Concentration of nonregulated DBPs in plant effluents; disinfection byproducts formed in the presence of bromide. percentage of TOX accounted for in plant effluents in U.S. Environ. Sci. Technol. 1999, 33 (19), 3378-3383. occurrence study (on a median basis); percentage of TOCl (16) Plewa, M. J.; Wagner, E. D.; Jazwierska, P.; Richardson, S. D.; Chen, P. H.; McKague, A. B. Halonitromethane drinking water accounted for in plant effluents; formation of HAAs after disinfection byproducts: chemical characterization and mam- chlorine dioxide and chlorine disinfection at two plants; EI malian cell cytotoxicity and genotoxicity. Environ. Sci. Technol. mass spectrum for DBP identified as 3,3-dibromo-4-oxo- 2004, 38 (1), 62-68.

7184 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 23, 2006 (17) Xie, Y.; Reckhow, D. A. Hydrolysis and dehalogenation of 217th National Meeting of the American Chemical Society, trihaloacetaldehydes. In Disinfection By-Products in Water Anaheim, CA, March 1999. Treatment: The Chemistry of Their Formation and Control; (35) Zhang, X.; Echigo, S.; Minear, R. A.; Plewa, M. J. Characterization Minear, R. A., Amy, G. L., Eds.; CRC Press/Lewis Publishers: and comparison of disinfection by-products of four major Boca Raton, FL, 1996; pp 283-291. disinfectants. In Natural Organic Matter and Disinfection By- (18) Thomas, R. F.; Weisner, M. J.; Brass, H. J. The fifth trihalo- Products: Characterization and Control in Drinking Water; methane: dichloroiodomethane, its stability and occurrences Barrett, S. E., Krasner, S. W., Amy, G. L., Eds.; ACS Symposium in chlorinated drinking water. In Water Chlorination: Envi- Series 761; American Chemical Society: Washington, DC, 2000; ronmental Impact and Health Effects; Ann Arbor Science: Ann pp 299-314. Arbor, MI, 1980; Vol. 3, p 161-168. (36) Amy, G. L.; Tan, L.; Davis, M. K. The effects of ozonation and (19) Hansson, R. C.; Henderson, M. J.; Jack, R.; Taylor, R. D. Iodoform activated carbon adsorption on trihalomethane speciation. taste complaints in chloramination. Water Res. 1987, 21 (10), Water Res. 1991, 25 (2), 191-202. 1265-1271. (37) Colclough, C. A.; Johnson, J. D.; Christman, R. F.; Millington, (20) Cancho, B.; Ventura, M.; Galceran, M.; Diaz, A.; Ricart, S. D. S. Organic reaction products of chlorine dioxide and natural Determination, synthesis and survey of iodinated trihalo- aquatic fulvic acids. In WATER CHLORINATION: Environmental methanes in water treatment processes. Water Res. 2000, 34 Impact and Health Effects; Jolley, R. L., et al., Eds.; Ann Arbor (13), 3380-3390. Science: Ann Arbor, MI, 1983; Vol. 4, pp 219-229. (21) Bichsel, Y.; von Gunten, U. Formation of iodo-trihalomethanes (38) Richardson, S. D.; Thruston, A. D., Jr.; Collette, T. W.; Patterson, during disinfection and oxidation of iodide-containing waters. K. S.; Lykins, B. W., Jr.; Majetich, G.; Zhang, Y. Multispec- Environ. Sci. Technol. 2000, 34 (13), 2784-2791. tral identification of chlorine dioxide disinfection byproducts (22) Woo, Y.-T.; Lai, D.; McLain, J. L.; Manibusan, M. K.; Dellarco, in drinking water. Environ. Sci. Technol. 1994, 28, 8 (4), 592- V. Use of mechanism-based structure-activity relationships 599. analysis in carcinogenic potential ranking for drinking water (39) Richardson, S. D.; Thruston, A. D., Jr.; Rav-Acha, C.; Groisman, disinfection by-products. Environ. Health Perspect. 2002, 110 L.; Popilevsky, I.; Juraev, O.; Glezer, V.; McKague, A. B.; Plewa, (suppl. 1), 75-87. M. J.; Wagner, E. D. Tribromopyrrole, brominated acids, and (23) Weinberg, H. S.; Krasner, S. W.; Richardson, S. D.; Thruston, A. other disinfection byproducts produced by disinfection of D., Jr. The Occurrence of Disinfection By-Products (DBPs) of drinking water rich in bromide. Environ. Sci. Technol. 2003, 37 Health Concern in Drinking Water: Results of a Nation- (17), 3782-3793. wide DBP Occurrence Study; EPA/600/R-02/068; U.S. EPA: (40) Richardson, S. D.; Thruston, A. D., Jr.; Krasner, S. W.; Weinberg, Athens, GA, 2002. www.epa.gov/Athens/publications/reports/ H. S.; Miltner, R. J.; Narotsky, M. G.; Simmons, J. E. Integrated EPA_600_R02_068.pdf. disinfection byproducts mixtures research: comprehensive (24) Kundu, B.; Richardson, S. D.; Granville, C. A.; Shaughnessy, D. characterization of water concentrates prepared from postchlo- T.; Hanley, N. M.; Swartz, P. D.; Richard, A. M.; DeMarini, D. rinated and preozonated/postchlorinated drinking water. J. M. Comparative mutagenicity of halomethanes and haloni- Toxicol. Environ. Health, submitted for publication. tromethanes in Salmonella TA100 structure-activity analysis (41) Plewa, M. J.; Wagner, E. D.; Jazwierska, P.; Richardson, S. D.; and mutation spectra. Mutat. Res. 2004, 554, 335-350. Chen, P. H.; McKague, A. B. Chemical and biological charac- (25) Plewa, M. J.; Kargalioglu, Y.; Vankerk, D.; Minear, R. A.; Wagner, terization of newly discovered iodoacid drinking water disin- E. D. Development of a quantitative comparative cytotoxocity fection byproducts. Environ. Sci. Technol. 2004, 38 (18), 4713- and genotoxicity assay for environmental hazardous chemicals. 4722. Water Sci. Technol. 2000, 42 (7-8), 109-116. (42) Hunter, E. S., III.; Tugman, J. A. Inhibitors of glycolytic (26) Krasner, S. W.; Pastor, S.; Chinn, R.; Sclimenti, M. J.; Weinberg, metabolism affect neurulation-staged mouse conceptuses in H. S.; Richardson, S. D.; Thruston, A. D., Jr. The occurrence of vitro. Teratology 1995, 52 (6), 317-323. a new generation of DBPs (beyond the ICR), Proceedings of the (43) Cemeli, E.; Wagner, E. D.; Anderson, D.; Richardson, S. D.; Plewa, 2001 AWWA WQTC; AWWA: Denver, CO, 2001. M. J. Modulation of the cytotoxicity and genotoxicity of the (27) Sclimenti, M. J.; Krasner, S. W.; Richardson, S. D. The deter- DBP iodoacetic acid by suppressors of oxidative stress. Environ. mination of DBPs using a solid-phase microextraction (SPME)- Sci. Technol. 2006, 40 (6), 1878-1883. GC/ECD technique, Proceedings of the 2002 AWWA WQTC; (44) McKnight, A.; Reckhow, D. A. Reactions of ozonation by-products AWWA: Denver, CO, 2002. with chlorine and chloramines, Proceedings of the 1992 AWWA (28) Onstad, G. D.; Weinberg, H. S. Evaluation of the stability and Annual Conference; AWWA: Denver, CO, 1992; pp 399-409. analysis of halogenated furanones in disinfected drinking waters. (45) Hoigne´, J.; Bader, H. The formation of trichloronitromethane Anal. Chim. Acta 2005, 534 (2), 281-292. (chloropicrin) and chloroform in a combined ozonation/ (29) Munch, D. J.; Munch, J. W.; Pawlecki, A. M. Method 552.2: chlorination treatment of drinking water. Water Res. 1988, 22 Determination of haloacetic acids and dalapon in drinking water (3), 313-319. by liquid-liquid extraction, derivatization and gas chroma- (46) Symons, J. M.; Krasner, S. W.; Simms, L. A.; Sclimenti, M. J. tography with electron capture detection. In Methods for the Measurement of THM and precursor concentrations revisited: Determination of Organic Compounds in Drinking Water, the effect of bromide ion. J. Am. Water Works Assoc. 1993, 85 Supplement III; EPA/600/R-95/131; U.S. EPA: Cincinnati, OH, (1), 51-62. 1995. (47) Metropolitan Water District of Southern California; James M. (30) Brophy, K. S.; Weinberg, H. S.; Singer; P. C. Quantification of Montgomery Consulting Engineers Inc. Pilot-Scale Evaluation nine haloacetic acids using gas chromatography with electron of Ozone and PEROXONE; AWWARF: Denver, CO, 1991. capture detection. In Natural Organic Matter and Disinfection (48) Simpson, K. L.; Hayes, K. P. Drinking water disinfection by- By-Products: Characterization and Control in Drinking Water; products: an Australian perspective. Water Res. 1998, 32 (5), Barrett, S. E., Krasner, S. W., Amy, G. L., Eds.; ACS Symposium 1522-1528. Series 761; American Chemical Society: Washington, DC, 2000; (49) Samsonov, D. P.; Pasynkova, E. M.; Bornovalova, G. V. Deter- pp 343-355. mination of 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-fura- (31) Standard Methods for the Examination of Water and Wastewater, none (toxicant MX) in drinking water by chromatography- 20th ed.; APHA, AWWA, and WEF: Washington, DC, 1998. tandem mass spectrometry. J. Anal. Chem. 2002, 57 (6), 513- (32) Echigo, S.; Zhang, X.; Minear, R. A.; Plewa, M. J. Differentiation 517. of total organic brominated and chlorinated compounds in total (50) Exner, J. H.; Burk, G. A.; Kyriacou, D. Rates and products of organic halide measurement: a new approach with an ion- decomposition of 2,2-dibromo-3-nitrilopropionamide. J. Agric. chromatographic technique. In Natural Organic Matter and Food Chem. 1973, 21, 838-842. Disinfection By-Products: Characterization and Control in (51) Wagner, E. D.; Muellner, M. G.; Richardson, S. D.; Plewa, M. J. Drinking Water; Barrett, S. E., Krasner, S. W., Amy, G. L., Eds.; Mammalian cell cytotoxicity and genotoxicity of new drinking ACS Symposium Series 761; American Chemical Society: water disinfection by-products. Abstracts of the 9th International Washington, DC, 2000; pp 330-342. Conference on Environmental Mutagens. Mutat. Res. 2005, 577 (33) Richardson, S. D.; Thruston, A. D., Jr.; Caughran, T. V.; Chen, (supp. 1), e1-e256. P. H.; Collette, T. W.; Floyd, T. L.; Schenck, K. M.; Lykins, B. W., Jr.; Sun, G.-R.; Majetich, G. Identification of new ozone disinfection byproducts in drinking water. Environ. Sci. Technol. Received for review February 15, 2006. Revised manuscript 1999, 33, 3 (19), 3368-3377. received June 9, 2006. Accepted June 13, 2006. (34) Khiari, D.; Ventura, F.; Chinn, R.; Barrett, S. The occurrence of iodinated trihalomethanes in drinking water. Presented at the ES060353J

VOL. 40, NO. 23, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7185